Latest Updates:
Page Index Toggle Pages: 1 ... 5 6 [7] 
Topic Tools
Very Hot Topic (More than 25 Replies) Can Fritz estimate my Elo rating? Yes. (Read 30960 times)
ErictheRed
God Member
*****
Online


USCF National Master

Posts: 2484
Location: USA
Joined: 10/02/05
Re: Can Fritz estimate my Elo rating? Yes.
Reply #11 - 06/03/14 at 04:53:51
Post Tools
RdC wrote on 06/02/14 at 23:30:29:
...when a tournament director becomes suspicious for one reason or another and wants to take action, Regan is the first man to get a call.



Seems like people are a little too paranoid, in my opinion.  I'm glad that someone is doing something about this potential problem, personally, so long as he doesn't have the authority of an Inquisitor.
  
Back to top
 
IP Logged
 
RdC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 865
Joined: 05/17/08
Re: Can Fritz estimate my Elo rating? Yes.
Reply #10 - 06/02/14 at 23:30:29
Post Tools
ErictheRed wrote on 06/02/14 at 22:52:35:
Ken is looking for statistical evidence, not starting a witch hunt.


There's an article on the USCF website at
http://www.uschess.org/content/view/12677/763/

In this I read

The current anti-cheating regulations of the world chess federation (FIDE) are too outdated to include guidance about disciplining illegal computer assistance, so Regan himself monitors most major events in real-time, including open events, and when a tournament director becomes suspicious for one reason or another and wants to take action, Regan is the first man to get a call.

That reads to me that he's appointed himself as the witch-finder.

Certainly you can test whether a player's moves match those of an engine. You might infer that the player consulted a computer engine. That's as far as you can go in the absence of physical evidence or observation from the scene as there's a reasonable doubt that it was all just preparation. It is not cheating to consult a computer engine before the game has started. I see no recognition of that in the Chess Life article.



  
Back to top
 
IP Logged
 
ErictheRed
God Member
*****
Online


USCF National Master

Posts: 2484
Location: USA
Joined: 10/02/05
Re: Can Fritz estimate my Elo rating? Yes.
Reply #9 - 06/02/14 at 22:52:35
Post Tools
This is a bit silly, we've had this argument before (mostly in the Ivanov thread).  What does "valid accusation" even mean?  It's not a guilty sentence, is it?  Perhaps "probable cause" is a better term.

Anyhow as with everything in life outside of logic and mathematics, we accumulate evidence to support our assertions.  I'd think that what Ken Regan describes could constitute evidence of cheating.  Enough to warrant a guilty verdict?  Depends on the case, I suppose, but lacking any other evidence perhaps not.  Ken is looking for statistical evidence, not starting a witch hunt.
  
Back to top
 
IP Logged
 
RdC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 865
Joined: 05/17/08
Re: Can Fritz estimate my Elo rating? Yes.
Reply #8 - 06/02/14 at 22:34:49
Post Tools
Bibs wrote on 06/02/14 at 12:24:41:
I think RdC is misunderstanding Regan's work, or deliberately misrepresenting.
It appears to me like Regan is taking time out of his professional work to try to apply the skill set he has to help root out cheaters in the chess world. I believe that such considered efforts should be welcomed. Respect.


I don't believe I am misrepresenting him. He has expressed the belief that it is possible to make a valid accusation of cheating in an OTB tournament purely on the basis of analysis of a player's games. I dispute that belief and believe you need the additional evidence of how the computer engine was consulted.

http://www.chessprofessionals.org/content/draft-fideacp-anti-cheating-proposal is the ACP/FIDE proposal although FIDE have yet to endorse it.


The Committee recommends the implementation of a FIDE Internet-based Game Screening Tool for pre-scanning games and identifying potential instances of cheating, together with the adoption of a full-testing procedure in cases of complaints. Together they shall meet the highest academic and judicial standards, in that they have been subject to publication and peer review, have a limited and documented error rate, have undergone vast empirical testing, are continuously maintained, and are generally accepted by the scientific community. Once in place, the Internet-based Game Screening Tool will be accessible to arbiters and chess officials and will be a useful instrument to prevent fraud, while the full test procedure will adhere to greater privacy as managed by FIDE and ACC.



  
Back to top
 
IP Logged
 
Smyslov_Fan
YaBB Moderator
Correspondence fan
*****
Offline


Progress depends on the
unreasonable man. ~GBS

Posts: 6902
Joined: 06/15/05
Re: Can Fritz estimate my Elo rating? Yes.
Reply #7 - 06/02/14 at 16:44:06
Post Tools
The article is a fascinating read.

We don't need to guess what Regan thinks of computers compared to "perfect" play. He states that perfect play is probably about 3600 elo and today's engines are around 3200-3300 strength. He states that computers are getting closer to perfection every year, gaining about 20 elo per year.
  
Back to top
 
IP Logged
 
Stigma
God Member
*****
Offline


There is a crack in everything.

Posts: 3107
Joined: 11/07/06
Gender: Male
Re: Can Fritz estimate my Elo rating? Yes.
Reply #6 - 06/02/14 at 15:38:24
Post Tools
I believe I had an opponent cheating against me once. He was much lower rated, yet outplayed me for most of the game while mostly leaving the table while I was thinking. When I got into serious time trouble and he could no longer leave the board inconspicuously, he was suddenly playing much worse and I won easily.

This may seem rather flimsy evidence, but the player in question was in fact caught a couple of years later and his accomplice (who was probably using the unsophisticated method of standing outside the playing hall with a smartphone running an engine...) banned from attending his tournaments.

What really astonished me was just how much suspicion and circumstancial evidence was needed before the federation eventually took action and treated the case. Nobody likes to accuse anyone of cheating; TDs and officials want to be really sure before they do that. Which makes me think quite a few people actually get away with it by cheating more cleverly (hidden device, not choosing typical "computer moves").

Opens and amateur tournaments typically can't afford any high-tech anti-cheating measures, and usually even allow phones and computers in the playing hall "as long as they're switched off". This may just be too naïve these days.
  

Improvement begins at the edge of your comfort zone. -Jonathan Rowson
Back to top
 
IP Logged
 
MartinC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 1982
Joined: 07/24/06
Re: Can Fritz estimate my Elo rating? Yes.
Reply #5 - 06/02/14 at 15:07:36
Post Tools
Its as much a measure against random witch hunts as actual cheating.
  
Back to top
 
IP Logged
 
Keano
God Member
*****
Offline


Money doesn't talk, it
swears.

Posts: 2891
Location: Toulouse
Joined: 05/25/05
Gender: Male
Re: Can Fritz estimate my Elo rating? Yes.
Reply #4 - 06/02/14 at 14:33:14
Post Tools
A human could cheat more cleverly by choosing the 3rd or 4th choice of the engine in most cases.

Analysing moves and comparing to engines will always be futile in my opinion. Luckily it is very very difficult to cheat at chess, and very few players are inclined to do it anyway. I play regularly and have never suspected anybody.

There are some high profile cases, but in these the level of sophistication involved has been mind boggling.

In short, cheating is not a big problem in chess despite what you read in chessbase or other places.
  
Back to top
 
IP Logged
 
ErictheRed
God Member
*****
Online


USCF National Master

Posts: 2484
Location: USA
Joined: 10/02/05
Re: Can Fritz estimate my Elo rating? Yes.
Reply #3 - 06/02/14 at 13:51:19
Post Tools
It seems to me that the assumption is that if humans use computer engines to cheat, their moves will more likely correlate with engine moves, not that engines know the truth about chess.

More info would be welcome.
  
Back to top
 
IP Logged
 
Bibs
God Member
*****
Offline



Posts: 2138
Joined: 10/24/06
Re: Can Fritz estimate my Elo rating? Yes.
Reply #2 - 06/02/14 at 12:24:41
Post Tools
RdC wrote on 06/02/14 at 09:42:55:
GeneM wrote on 06/02/14 at 06:44:19:
. .

However, Regan's main goal is to better detect chess players who cheat by sneaking a use of Fritz.



Or Stockfish, Houdini, Komodo etc.

The method is flawed because it is quite legal to use engine moves in Over the Board competitions, provide that is, you or someone else analysed them beforehand in the calm of their study.

By contrast on-line competitions played in the absence of witnesses should never be used for important events and ratings. It's not just computer engines you have to worry about, it's also the open book next to the screen or the stronger player standing behind suggesting moves.

There's a hidden assumption behind ideas like IPR that computer engines know the truth about chess. I don't think that's true, or at least not yet. A couple of points:-
There are still engine v engine contests which result in decisive results.
Anand commented within recent years that if he went back and rechecked engine recommendations from a few years ago, the evaluations had changed. This would be both because of deeper searching and refining of parameters by the programmers.


I think RdC is misunderstanding Regan's work, or deliberately misrepresenting.
It appears to me like Regan is taking time out of his professional work to try to apply the skill set he has to help root out cheaters in the chess world. I believe that such considered efforts should be welcomed. Respect.
1. Regan is/was a very reasonable player. Regarding stats,  he appears eminent in this regard. I am sure he is well aware of theory being generated by softwares. Why do you assume he is ignorant of this? Frankly, this seems a ridiculous claim.
2. He is not assuming chess 'truth' of softwares. Has he stated this? No. But there will be few who disagree re: strength superiority, and the matches by Regan are for human moves and for those of superior softwares.

Non-disclaimer - I have never met Regan, nor had any contact with him.
  
Back to top
 
IP Logged
 
RdC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 865
Joined: 05/17/08
Re: Can Fritz estimate my Elo rating? Yes.
Reply #1 - 06/02/14 at 09:42:55
Post Tools
GeneM wrote on 06/02/14 at 06:44:19:
. .

However, Regan's main goal is to better detect chess players who cheat by sneaking a use of Fritz.



Or Stockfish, Houdini, Komodo etc.

The method is flawed because it is quite legal to use engine moves in Over the Board competitions, provide that is, you or someone else analysed them beforehand in the calm of their study.

By contrast on-line competitions played in the absence of witnesses should never be used for important events and ratings. It's not just computer engines you have to worry about, it's also the open book next to the screen or the stronger player standing behind suggesting moves.

There's a hidden assumption behind ideas like IPR that computer engines know the truth about chess. I don't think that's true, or at least not yet. A couple of points:-
There are still engine v engine contests which result in decisive results.
Anand commented within recent years that if he went back and rechecked engine recommendations from a few years ago, the evaluations had changed. This would be both because of deeper searching and refining of parameters by the programmers.
  
Back to top
 
IP Logged
 
GeneM
Senior Member
****
Offline


Tournament winner gets
two fun filled knights!

Posts: 303
Location: near Seattle WA USA
Joined: 01/12/08
Can Fritz estimate my Elo rating? Yes.
06/02/14 at 06:44:19
Post Tools
. .
I cannot find the forum post presently, but some years ago I expressed surprise that Fritz did not yet have a feature whereby it could play against me and afterward tell me my estimated Elo rating.

Now Ken Regan has written a computer program that can do this (see article quotes below).
However, Regan's main goal is to better detect chess players who cheat by sneaking a use of Fritz.

In an unrealistically ideal world, a perfect mechanism for detecting cheating just by analyzing the moves would allow chess games played over the web, at long time controls, to be formally rated by organizations like the USCF - even when no TD is at each player's home to chaperon.
Great reductions in travel costs, hotel costs, venue rental fees passed on to tournament entry fees, and poor quality downtime between at-venue morning to evening games - could all be reduced.

For long time control competition, a mix of web-based and in-person face-to-face games would be nicer than the present world of web-based play being limited to unrated games. The mix would also help ensure the web play rating was not substantially higher than the in-person rating.

------------

http://www.uschess.org/content/view/12677/763/
Article Title: "How To Catch A Chess Cheater: Ken Regan Finds Moves Out Of Mind"

Quotes from the article:
------------

Ken Regan takes a set of chess positions played by a single player—ideally 200 or more but his analysis can work with as few as 20—and treats each position like a question on a multiple-choice exam. The score on this exam translates to an Elo rating, a score Regan calls an Intrinsic Perfor­mance Rating (IPR).
...
In Regan’s algorithms it is the relative differences in move quality that matter, not the absolute differences.
...
Regan would like to oversee the conversion of his 35,000 lines of C++ code into a Windows-driven program or portable app.
...
The most notorious public cheating case to date has been that of the then-26-year-old Bulgarian Borislav Ivanov. ... Regan’s analysis found that Ivanov’s moves ... translates to the odds of him independently making these moves to less than one chance in five million.
...
inevitable false positives. In any large open tournament with at least a thousand non-cheating players, the chances are very high that at least one of those honest players will earn a ... suspicious value.
...
data shows that players make 60 percent to 90 percent more errors when half a pawn ahead or behind than when the game is even.
...
The greatest over-the-board practical problems are not always caused by the objectively best moves, and Regan’s metric can quantify this distinction.
...
statistician Jeff Sonas has been rating historical players, but Regan’s IPR is more objective. Sonas uses historical game results [whereas Regan assess individual moves].
...
since Regan’s method compares moves to a common standard (the engine), rather than the results of games, he can objectively relate player abilities across eras. What he found was that rating inflation does not exist. ...
Thus one may conclude that Hikaru Nakamura’s peak FIDE rating of 2789 beats Bobby Fischer’s peak of 2785.
...
But selective-move cheaters would be doing it on critical moves, and Regan has untested tricks for these cases.


(End)
. .
  

GeneM , CastleLong.com , FRC-chess960
Back to top
WWW  
IP Logged
 
Page Index Toggle Pages: 1 ... 5 6 [7] 
Topic Tools
Bookmarks: del.icio.us Digg Facebook Google Google+ Linked in reddit StumbleUpon Twitter Yahoo